Magnesium–calcium signalling in rat parotid acinar cells:effects of acetylcholine

Antonio Mata Æ Duarte Marques Æ Marıa A. Martınez-Burgos ÆJoao Silveira Æ Joana Marques Æ Maria F. Mesquita Æ Jose A. Pariente ÆGines M. Salido Æ Jaipaul Singh

Received: 2 July 2007 / Accepted: 23 August 2007 / Published online: 12 September 2007

� Springer Science+Business Media, LLC 2007

Abstract This study investigated the effects of extracel-

lular Mg2+ ([Mg2+]o) on basal and acetylcholine (ACh)-

evoked amylase secretion and intracellular free Ca2+

([Ca2+]i) in rat parotid acinar cells. In a medium containing

1.1 mM [Mg2+]o, ACh evoked significant increases in

amylase secretion and [Ca2+]i. Either low (0 mM) or ele-

vated (5 and 10 mM) [Mg2+]o attenuated ACh-evoked

responses. In a nominally Ca2+ free medium, elevated

[Mg2+]o attenuated basal and ACh-evoked amylase secre-

tion and [Ca2+]i. In parotid acinar cells incubated with

either 0, 1.1, 5 or 10 mM [Mg2+]o, ACh evoked a gradual

decrease in [Mg2+]i. These results indicate that the ACh-

evoked Mg2+ efflux is an active process since Mg2+ has to

move against its gradient. Either lidocaine, amiloride, N-

methyl-D-glucamine, quinidine, dinitrophenol or bumeta-

nide can elevate [Mg2+]i above basal level. In the presence

of these membrane transport inhibitors, ACh still evoked a

decrease in [Mg2+]i but the response was less pronounced

with either [Na+]o removal or in the presence of either

amiloride or quinidine. These results indicate marked

interactions between Ca2+ and Mg2+ signalling in parotid

acinar cells and that ACh-evoked Mg2+ transport was not

dependent upon [Na+]o.

Keywords Parotid � Amylase � Calcium � Magnesium �Acetylcholine � Lidocaine � Amiloride � Bumetanide

Introduction

Salivary gland secretory function is regulated by the two

neurotransmitters noradrenaline (NA) and acetylcholine

(ACh), which trigger mainly a sequence of cellular signal

transduction events resulting in intracellular cascades to

generate such intracellular messengers as Ca2+ and aden-

osine 3,5 cyclic monophosphate (cyclic AMP) which, in

turn, activate ion transport pathways, water and protein

secretion [1–4]. The precise cellular mechanisms by which

the intracellular messengers regulate salivary gland func-

tion are still not fully understood [5]. Knowledge in this

field can help to develop therapeutic strategies for resolu-

tion of some of the diseased states such as xerostomia,

mucositis and high caries index among others usually

associated with salivary gland dysfunctions.

On the other hand, Mg2+, whose biological importance

has become gradually recognised over the last century

[6, 7], is the second most abundant intra-cellular divalent

cation, exceeded only by K+ [8]. Mg2+ is involved in

numerous biological processes, including regulation of

enzyme and hormone secretion and several membrane ion

transport systems (e.g. ion channels, membrane ATPases,

Na+/K+/Cl–, Na+/H+, K+/Cl–) in epithelial secretory cells

[5, 6, 7, 9, 10–12]. Either a derangement in Mg2+

homeostasis (e.g. hypomagnesemia or reduction in dietary

A. Mata

Department of Oral Biology, Faculty of Dental Medicine,

University of Lisbon, Cidade Universitaria, Lisboa, Portugal

D. Marques � J. Silveira � J. Marques � M. F. Mesquita

Oral Biology Research Group, Instituto Superior de Ciencias da

Saude Egas Moniz, Monte Caparica, Portugal

M. A. Martınez-Burgos � J. A. Pariente � G. M. Salido

Department of Physiology, Cell Physiology Research Group,

University of Extremadura, Caceres, Spain

J. Singh (&)

Department of Biological, Forensic Sciences, University

of Central Lancashire,

Preston PR1 2 HE Lancashire, England, UK

e-mail: jsingh3@uclan.ac.uk

123

Mol Cell Biochem (2008) 307:193–207

DOI 10.1007/s11010-007-9599-y

intake of Mg2+) can lead to a number of pathophysiological

conditions including cardiovascular diseases [13], peptic

ulcers [14] and diabetes [15]. Mg2+ is believed to exert

some of its protective effects by regulatory cellular Ca2+

transport [7, 11]. Indeed, Mg2+ has been described as nat-

ural antagonist for Ca2+ [6]. The intracellular free calcium

concentration ([Ca2+]i) in control condition is normally

kept at low levels compared to the intracellular magnesium

concentration ([Mg2+]i). The reason is that Ca2+ transport is

tightly controlled by the plasma membrane and by intra-

cellular organelles [5]. Upon stimulation, [Ca2+]i is

elevated whereas [Mg2+]i is decreased [11]. These reci-

procal changes suggest some interaction between these two

divalent cations during cellular regulation.

The relationship between Ca2+ and Mg2+ signalling has

been studied extensively in the pancreas [7, 11], parietal

cells, red blood cells and the heart [1, 13, 14, 16]. Simi-

larly, other studies have investigated the relationship

between Na+ and Mg2+ signalling in salivary sublingual

acinar cells [17, 18]. However, in the salivary parotid

glands very little is known about the regulatory effects of

Mg2+ and its interactions with Ca2+ either during basal

conditions or during secretagogue-evoked secretory

responses compared to the sublingual gland. The parotid

gland plays a major role in the secretion of amylase and

fluid for the digestion of food [1, 2]. Therefore, this study

investigated the biological role and transport of Mg2+ in the

parotid gland and its relationship with intracellular Ca2+

homeostasis during ACh-induced amylase secretion.

Materials and methods

Materials

Magfura-2 (AM) and fura-2 (AM) were from Molecular

Probes (Eugene, Oregon, USA). All other reagents were of

the highest purity available and purchased from Sigma

(Madrid, Spain) unless indicated otherwise.

Experimental procedure

Institutional Ethics Committees approved all procedures.

Adult male and female CD strain of Sprague–Dawley rats

weighing 200–300 g were used in this study. Animals were

humanely killed by blow to the head followed by cervical

dislocation. An incision was made in upper part of the neck

and the parotid glands were located and quickly removed

and placed into a modified Krebs–Heinselet (KH) solution

containing (mM): NaCl, 103; KCl, 4.76; CaCl2, 2.56;

MgCl2, 0, 1.1, 5 or 10; NaHCO3, 25; NaH2PO4, 1.15;

D-Glucose, 1.8; sodium pyruvate, 4.9; sodium fumarate,

2.7 and sodium glutamate, 4.9. The solution was kept at pH

7.4 while being continuous gassed with a mixture of 95%

O2–5% CO2 and maintained at 37�C. A concentration of

1.1 mM Mg2+ was used as the normal Mg2+ concentration

outside the cells ([Mg2+]o) in this study since this is the

physiological plasma Mg2+ concentration and moreover,

we had previously used the same concentration to study the

effects of changes of [Mg2+]o on nerve-mediated and

secretagogue-evoked secretory responses in the exocrine

pancreas [9, 11, 12, 19]. In addition, both zero and an

elevated (5 and 10 mM) [Mg2+]o were used for comparison

in order to investigate the effects of [Mg2+]o.

Measurement of amylase release

The parotid glands were cut into small segments (10–

15 mg) and a total weight of about 100–150 mg was placed

in a Perspex flow chamber (volume 1 ml) and superfused

with oxygenated KH solution at 37�C at a flow rate of

1.8 ml min–1. After tissue stabilisation, the amylase con-

centration (in Units (U) ml–1) in the effluent from the

chamber was measured continuously using an on-line

fluorimetric method [20] that depends essentially upon the

liberation of dialyzable fluorogenic products from amylo-

pectin anthranilate used as a substrate. a-Amylase (Sigma

type II-A) was used as a standard for calibration. ACh was

added directly to the superfusing solution to give concen-

trations of 10–8–10–5 M. The tissue was stimulated with

ACh for 6 min. At the end of each experiment, the tissue

was removed from the chamber, blotted dry and weighed,

and amylase release was expressed in U ml–1 (100 mg

tissue)–1. For superfusion with either different Mg2+ con-

centrations or with a nominally Ca2+-free medium, either

MgCl2 or CaCl2 was replaced with NaCl in order to

maintain a constant osmolarity in the KH solution. Fur-

thermore, in those experiments in which external Ca2+

([Ca2+]o) was depleted, EGTA (1 mM) was added thus

making the solution nominally Ca2+-free.

Preparation of parotid acinar cells

The parotid glands were dissociated into acinar cells with

collagenase in two stages after a period of 75 min by

established methods [9, 21]. Briefly, the parotid was

incubated in the presence of collagenase for 10 min at

37�C. This enzymatic digestion of the tissue was followed

by gently pipetting the cell suspension through tips of

decreasing diameter for mechanical dissociation of the

acinar cells. After centrifugation, cells were resuspended in

a Hanks balanced salt solution (HBSS) HEPES-buffered

saline containing (in mM): 10 HEPES, 140 NaCl, 4.7 KCl;

194 Mol Cell Biochem (2008) 307:193–207

123

1.1 MgCl2; 10 glucose and 1.8 CaCl2; pH 7.40. With this

isolation procedure, single cells as well as small clusters

consisting up to five cells were obtained. Cell viability

monitored with trypan blue was greater than 95% [22].

Cell loading and measurement of [Ca2+]i and [Mg2+]i

The cell suspension was incubated with either 2 lM fura-

2-AM (for [Ca2+]i measurement) or magfura-2-AM (for

[Mg2+]i measurement) in the presence of 0.025% pluronic

acid at room temperature for 25 min using an established

method [23–25]. Following loading, the cells were centri-

fuged for 3 min at 50 · g and resuspended in fresh HBSS

and used within 2–3 h following loading. For determina-

tion of fluorescence, a volume of 200–300 ll of cell

suspension was placed on a poly-D-lysine (20 lg ml–1)

coated thin glass cover slip attached to a Perspex perfusion

chamber, which was continuously perfused with HBSS

(approximately at a rate of 1.5 ml min–1) at room tempera-

ture [23, 25]. The perfusion chamber was placed on the stage

of an inverted fluorescence microscope (Nikon Diaphot

200). The cells (50 individual cells were chosen) were

alternatively excited at 340 and 380 nm by computer-

controlled filter wheel (Lambda-2, Sutter Instruments) and

the emitted images ([510 nm) were captured by a high speed

cooled digital CCD camera (C-4880–81, Hamamatsu Pho-

tonics, Shizuoka, Japan), and recorded using an appropriate

software (Argus-HiSCa, Hamamatsu Photonics, Shizuoka,

Japan). All values were measured in arbitrary fluorescence

ratio units (F340/F380). ACh was added directly to the

superfusing medium at a concentration of 10–5 M and

extracellular Mg2+ and Ca2+ were modified appropriately. In

the experiments where extracellular Ca2+ was removed,

1 mM EGTA was added to the superfusing medium in order

to obtain a real free Ca2+ solution.

Statistical analysis

All data provided were expressed as means ± standard

error of the mean (SEM). Data were compared by ANOVA

and post Hoc tests and only values with P \ 0.05 were

accepted as significant.

Results

Figure 1 shows the effect of varying the concentrations

(10–8–10–5 M) of ACh on amylase output from superfused

parotid segments in normal 2.56 mM [Ca2+]o but during

the perturbation of [Mg2+]o. Original chart recordings of

the response for 10–5 M ACh in 0, 1.1, 5 and 10 mM

[Mg2+]o are shown in (A) and the mean (±SEM) amylase

output above basal level are shown in Fig. 1B for all

concentrations of ACh. The results show that ACh can

evoke a dose dependent increase in amylase output at all

[Mg2+]o with maximal effect in 1.1 mM [Mg2+]o. Both

zero and elevated [Mg2+]o attenuated the ACh-evoked

secretory responses.

Fig. 1 ACh-evoked amylase

secretion in isolated parotid

gland segments. (A) Original

chart recordings showing the

effect of 10–5 M ACh on

amylase secretion from isolated

superfused rat parotid gland

segments in (a) 0 mM, (b)

1.1 mM, (c) 5 mM and (d)

10 mM [Mg2+]o. Traces are

typical of 8–15 such

experiments taken from the

same number of animals.

Vertical and horizontal bars

show the concentration of

amylase output (U ml–1

(100 mg of tissue)–1) and time

(in min), respectively. (B)

Histograms showing the mean

(±SEM) amylase output above

basal level during stimulation

with 10–8–10–5 M ACh in 0,

1.1, 5 and 10 mM [Mg2+]o,

(n = 8–15), *P \ 0.01

compared with values obtained

in normal 1.1 mM [Mg2+]o

Mol Cell Biochem (2008) 307:193–207 195

123

When basal amylase secretion was stabilised, [Ca2+]o was

totally removed from perfusing medium and 1 mM EGTA

was added (Fig. 2A (b)), parotid segments were then stim-

ulated with 10–5 M ACh (Fig. 2A (c)). After peak recovery,

segments were again perfused with an extracellular

medium containing 2.56 mM [Ca2+]o, in the absence of the

secretagogue, (Fig. 2A (d)). These experiments were

repeated in the presence of different [Mg2+]o namely: 0,

1.1, 5 and 10 mM. Figure 2B shows the mean (±SEM)

amylase secretion for basal in 2.56 mM [Ca2+]i, and in a

Fig. 2 ACh-evoked amylase secretion from parotid gland segments

in the absence and presence of [Ca2+]o. (A) Original chart recordings

showing time course of amylase secretion from superfused parotid

segments in (a) normal (2.56 mM) [Ca2+]o, (b) a nominally free

[Ca2+]o containing 1 mM EGTA in the absence (c) presence of

10–5 M ACh in zero [Ca2+]o + 1 mM EGTA and (d) following

reperfusion with normal (2.56 mM) [Ca2+]0 alone during perturbation

of [Mg2+]0. Traces are typical of 8–10 such experiments taken from the

same number of animals. (B) Bar charts showing mean (±SEM) basal

amylase output in normal (2.56 mM) and in a nominally free [Ca2+]o

containing 1 mM EGTA during perturbation of [Mg2+]0, n = 8–10.

(C) Bar charts showing mean (±SEM) amylase output above basal

level from superfused parotid gland segments either during ACh

(10–5 M) application in a nominally free [Ca2+]o KH containing 1 mM

EGTA or following reperfusion with normal (2.56 mM) [Ca2+]o

during perturbation of [Mg2+]o, (n = 8–10). In Fig. 2B, P \ 0.01 for

(c) and (d) compared to (a) and (b) which are not significantly different

from one another. Similarly, P \ 0.01 for (f), (g) and (h) compared to

(e). In Fig. 2C, P \ 0.01 for (b), (c) and (d) compared to (a). Similarly,

P \ 0.01 for (f), (g) and (h) compared to (e)

196 Mol Cell Biochem (2008) 307:193–207

123

nominally free [Ca2+]o + 1 mM EGTA. Figure 2C shows

the ACh (10–5 M)-evoked mean (±SEM) amylase above

basal peak, in the absence of [Ca2+]o and re-introducing

normal (2.56 mM) [Ca2+]o in the presence of 0, 1.1, 5 and

10 mM [Mg2+]o. The results show that 5 and 10 mM

[Mg2+]o can significantly (P \ 0.05) attenuate basal (in the

presence and absence of [Ca2+]o), ACh-evoked amylase

output in the absence of [Ca2+]o and the secretory

response elicited by re-introducing normal (2.56 mM)

[Ca2+]o. In contrast, in zero [Mg2+]o and in a nominally

free [Ca2+]o + 1 mM EGTA either basal, ACh-evoked or

re-introducing normal [Ca2+]o resulted in large and sig-

nificant increases in amylase secretion compared to the

effects of elevated [Mg2+]o. These results suggest that

[Ca2+]o is required to maintain a high basal and secreta-

gogue-evoked amylase secretion and elevated Mg2+ seems

to regulate Ca2+ influx into the cell and possible its release

from the internal stores.

The next logical step in this study was to investigate the

effects of perturbation of [Mg2+] on basal and secreta-

gogue-evoked cellular Ca2+ homeostasis. Since 10–5 M

ACh elicited maximal secretory response and this con-

centration can mobilise [Ca2+]i effectively [9], then it was

decided to use this concentration to study cellular Ca2+

transport in this study. Figure 3 shows basal [Ca2+]i in

normal [Ca2+]o and in a nominally free [Ca2+]o HEPES

solution containing 1 mM EGTA during perturbation of

[Mg2+]o. All the results for [Ca2+]i are expressed in ratio

units [26]. The results show that elevated [Mg2+]o can

attenuate basal [Ca2+]i in parotid acinar cells either in

normal or in a nominally free [Ca2+]o HEPES solution

containing 1 mM EGTA. The response was much more

pronounced in a nominally free [Ca2+]o. However, in zero

[Mg2+]o, [Ca2+]i remains more or less the same in normal

[Ca2+]o compared to much reduced responses (P \ 0.05) in

5 mM and 10 mM [Mg2+]o. Taken together, the results

indicate that [Ca2+]o is required to maintain a high basal

level [Ca2+]i.

Figure 4A shows original time course traces of basal

and ACh (10–5 M)-evoked [Ca2+]i in normal [Ca2+]o and

during perturbation of [Mg2+]o. The results show that in

normal [Ca2+]o, ACh evoked a marked transient increase in

[Ca2+]i comprising an initial rise (peak) followed by a

gradual decrease which subsequently levels off to a plateau

phase (plateau) above basal level. The mean (±SEM) peak

and plateau increases in [Ca2+]i following stimulation of

fura-2-loaded parotid acinar cells with 10–5 M ACh in

different concentrations (0, 1.1, 5 and 10 mM) of [Mg2+]o

are shown in Fig. 4B. The results show that both zero and

elevated (5–10 mM) [Mg2+]o can significantly (P \ 0.05)

attenuate both the peak and the plateau phase of the Ca2+

transient evoked by ACh. The effect of 5 and 10 mM was

much more pronounced compared to zero [Mg2+]o. Since

both low (0 mM) and elevated [Mg2+]o can attenuate ACh-

evoked Ca2+ mobilisation, it was decided to investigate the

effect of perturbation of [Mg2+]o on Ca2+ mobilisation in a

nominally low [Ca2+]o HEPES containing 1 mM EGTA.

The rationale was to ascertain whether Mg2+ could regulate

ACh-evoked Ca2+ release from intracellular stores and

regulate its influx from the extracellular medium into the

cell.

Figure 5A shows original traces of [Ca2+]i in response to

10–5 M ACh in 0, 1.1, 5 and 10 mM [Mg2+]o in zero

[Ca2+]o. The mean (±SEM) peak and plateau responses are

shown in Fig. 5B. The results show that only elevated

(5 and 10 mM) [Mg2+]o can significantly (P \ 0.05)

attenuate the ACh–evoked increase in the initial peak phase

of the Ca2+ transient. In the presence of zero [Mg2+]o, the

initial ACh-evoked peak phase of the Ca2+ transient was

slightly larger compared to the responses obtained in

1.1 mM [Mg2+]o. Moreover, the results show that in the

absence of [Ca2+]o, the ACh-evoked plateau phase of the

Ca2+ transient was almost completely abolished suggesting

that extracellular Ca2+ is required to maintain the plateau

phase. Like the experiments on amylase secretion, these

results clearly indicate that elevated [Mg2+]o is antagonis-

ing the mobilisation of both basal and ACh-evoked cellular

Ca2+ mobilisation (Ca2+ release from intracellular stores

and Ca2+ influx into the cell). The results show that stim-

ulation of parotid acinar cells with ACh in the absence of

[Ca2+]o elicited a small, but rapid transient rise in [Ca2+]i

which rapidly returned to basal values. This type of signal

is described as Ca2+ exclusively exiting from intracellular

stores. After peak recovery, if cells are again perfused with

a medium containing normal [Ca2+]o a plateau shaped

rebound in [Ca2+]i is observed and this is described as

Fig. 3 Basal [Ca2+]i in fura-2-loaded parotid acinar cells. Mean

(±SEM) basal [Ca2+]i in fura-2-loaded parotid acinar cells in either

normal 1.8 or 0 mM [Ca2+]0 containing 1 mM EGTA during

perturbation of [Mg2+]o. Values are expressed as ratio units,

n = 15–20 cells taken from 8 to 10 rats. P \ 0.01 for (c) and (d)

compared to (a) and (b). Similarly, P \ 0.01 for (f), (g) and (h)

compared to (e)

Mol Cell Biochem (2008) 307:193–207 197

123

capacitative calcium entry (CCE), corresponding to Ca2+

entering the cell from the extracellular side [26, 27]. CCE

has been reported as the driving force for prolonged fluid

secretion in salivary glands [5]. Therefore, an investigation

was undertaken to determine the effects of a perturbation of

[Mg2+]o (0, 1.1, 5 and 10 mM) on CCE in this study.

Figure 6 shows original time course chart recordings of

the protocol to measure CCE following perturbation of

[Mg2+]o. Cells were superfused with a nominally

free [Ca2+]o HEPES solution containing 1 mM EGTA for

200–300 s and thereafter, stimulated with 10–5 M ACh

in the presence of (A) 0 mM, (B), 1.1 mM, (C) 5 mM and

(d) 10 mM [Mg2+]o. Following the peak and plateau

phases, both ACh and the nominally free [Ca2+]o medium

containing 1 mM EGTA were replaced with normal

HEPES solution containing 1.8 mM [Ca2+]o in the pres-

ence of low and elevated [Mg2+]o. Figure 7A shows the

time (downward arrows) it takes for CCE to be activated

following perfusion of fura-2-loaded acinar cells

with 1.8 mM [Ca2+]o HEPES solution in different (0, 1.1, 5

and 10 mM) [Mg2+]o. The results show that in a normal

(1.1 mM) [Mg2+]o, CCE is (P \ 0,01) rapidly activated

compared to delayed observations in zero and elevated

[Mg2+]o. These effects were also associated with similar

time course in amylase secretion (see Fig. 2A for com-

parison). Figure 7B shows the maximal increases in [Ca2+]i

following the re-introduction of 1.8 mM Ca2+ HEPES to

fura-2-loaded acinar cells at 100 s following activation of

CCE in different [Mg2+]o. The results showed that in zero

[Mg2+]o, there was a larger and sustained elevation in

[Ca2+]i after CCE activation. The effects of 5 and 10 mM

[Mg2+]o on [Ca2+]i were much smaller (P \ 0.01) com-

pared to the responses obtained in zero [Mg2+]o. These

results indicate that [Mg2+]o is regulating Ca2+ entry into

parotid acinar cells and the level of Ca2+ influx is also

dependent upon the concentration of [Mg2+]o. Low [Mg2+]o

Fig. 4 ACh-evoked increases

in [Ca2+]i in fura-2-loaded

single parotid acinar cells.

(A) Original chart recordings

showing the effect of

perturbation ((a) 0 mM, (b)

1.1 mM, (c) 5 mM and (d)

10 mM) of [Mg2+]o on ACh

(10–5 M)-evoked changes in

[Ca2+]i in fura-2-loaded single

parotid acinar cells during

perfusion with a HEPES

solution containing 1.8 mM

[Ca2+]o. Traces are typical of

15–20 cells, taken from 6 to 8

different animals. (B) Mean

(±SEM) changes in the peak and

plateau phases in the Ca2+

transient above basal level in

fura-2-loaded single parotid

acinar cells evoked by 10–5 M

ACh following perturbation of

[Mg2+]o. The cells were

perfused with a HEPES solution

containing 1.8 mM [Ca2+]o,

n = 15–20 cells taken from 8 to

10 animals. Note that

measurements were made

10–15 s (peak) and 2–3 min

(plateau) after ACh application.

P \ 0.01 for (a), (c) and (d)

compared to (b). Similarly,

P \ 0.01 for (e), (g) and (h)

compared to (f)

198 Mol Cell Biochem (2008) 307:193–207

123

seems to facilitate Ca2+ influx whereas elevated [Mg2+]o

has the opposite effect. Moreover, the response was much

more rapid in normal [Mg2+]o compared to other

concentrations.

Since perturbation of [Mg2+]o seems to have profound

effect on both basal and secretagogue-evoked amylase

secretion and [Ca2+]i, then it was pertinent to measure

[Mg2+]i during different [Mg2+]o either alone or following

perfusion of magfura-2-loaded single parotid acinar cells

with ACh. Figure 8A shows original time course chart

recordings of [Mg2+]i during basal condition and following

perfusion of parotid acinar cells with 10–5 M ACh in (a)

0 mM, (b) 1.1 mM, (c) 5 mM and (d) 10 mM [Mg2+]o. The

results show that ACh can elicit a marked time-dependent

decrease in [Mg2+]i reaching a plateau level after 10 min of

ACh application. Figure 8B shows mean (±SEM) basal

(prior to ACh application) and ACh-evoked steady-state

decrease (10 min after ACh application) in [Mg2+]i in 0, 1.1,

5 and 10 mM [Mg2+]0. The results show that basal [Mg2+]i

increased significantly (P \ 0.05) with elevated [Mg2+]0.

Moreover, the ACh-evoked decrease in [Mg2+]i was signif-

icantly (P \ 0.05) different from the respective control.

In another time course series of experiments, magfura

2-loaded acinar cells were perfused with different

Fig. 5 ACh-evoked changes in [Ca2+]i in fura-2-loaded single parotid

acinar cells during perturbation of [Mg2+]o. (A) Original chart

recordings showing the effect of perturbation ((a) 0 mM, (b) 1.1 mM,

(c) 5 mM and (d) 10 mM ) of [Mg2+]o on ACh (10–5 M)-evoked

changes in [Ca2+]i in fura-2-loaded single parotid acinar cells during

perfusion with a HEPES solution containing 0 mM [Ca2+]o and 1 mM

EGTA. Traces are typical of 15–20 cells taken from 6 to 8 different

animals. (B) Bar charts showing the mean (±SEM) changes in the

peak and plateau phases in the Ca2+ transient above basal level in

fura-2-loaded single parotid acinar cells evoked by 10–5 M ACh

following perturbation of [Mg2+]o. The cells were perfused with a

HEPES solution containing 0 mM [Ca2+]o and 1 mM EGTA, n =

15–20 cells taken from 8 to 10 animals. Note that measurements were

made 10–15 s (peak) and 2–3 min (plateau) after ACh application.

The results clearly show that in the absence of [Ca2+]o, elevated

[Mg2+]o attenuated the ACh-evoked peak response. However, the

ACh-evoked plateau phase of the Ca2+ transient was almost

completely abolished. P \ 0.05 for (b), (c) and (d) compared to (a)

Mol Cell Biochem (2008) 307:193–207 199

123

concentrations (0, 1.1, 5 and 10 mM) of [Mg2+]o for

15 min each with increased concentration followed by

stimulation with ACh (10–5 M) in the continuous presence

of 10 mM [Mg2+]o. An original chart recording of the time

course changes in [Mg2+]i is shown in Fig. 9A. The results

show that [Mg2+]i increases gradually following the

application of different [Mg2+]o. Perfusion of the cell with

10–5 M ACh in 10 mM [Mg2+]o resulted in a rapid

decrease in [Mg2+]i reaching a plateau within 2–3 min of

ACh application and remained at the same level even

during the removal of ACh. The mean (±SEM) increases in

[Mg2+]i and in different [Mg2+]o in the presence of ACh in

10 mM [Mg2+]o are shown in Fig. 9B. The results show

that perfusion of parotid acinar cells with different con-

centrations of [Mg2+]o resulted in a gradual rise in [Mg2+]i

and with a significant (P \ 0.01) increase in [Mg2+]i at

10 mM [Mg2+]o compared to the concentration obtained in

0 mM [Mg2+]o. In the presence of 10 mM [Mg2+]o, ACh

caused a marked and significantly (P \ 0.05) decrease in

[Mg2+]o suggesting that the ACh-induced decrease in

[Mg2+]i may be an energy-dependent process as Mg2+ has

to move against its gradient. Since ACh can elicit a marked

decrease in [Mg2+]i then it was necessary to understand the

mechanism of the ACh-induced Mg2+ transport.

Fig. 6 ACh-evoked changes in [Ca2+]i in fura-2-loaded parotid acinar

cells during perturbation of [Ca2+]o and [Mg2+]o. Time course of

changes in [Ca2+]i during perfusion of fura-2-loaded parotid acinar cells

in a nominally free [Ca2+]o containing 1 mM EGTA in (a) the absence

and (b) the presence of 10–5 M ACh and following re-introduction of

(c) 1.8 mM [Ca2+]o HEPES solution in (A) 0 mM, (B) 1.1 mM, (C)

5 mM and (D) 10 mM [Mg2+]o. Traces are typical of 10–12 such cells

taken from 5 to 8 rats. Traces have been used to analyse for data present

in Fig. 7

Fig. 7 CCE in fura-2-loaded parotid acinar cells during perturbation

of [Mg2+]o. (A) Bar charts showing the mean (±SEM) time

(in seconds) taken for CCE to be activated in 0, 1.1, 5 and 10 mM

[Mg2+]o following perfusion of fura-2-loaded parotid acinar cells with

1.8 mM [Ca2+]o HEPES solution, n = 10–12 cells taken from 5 to 8

animals. P \ 0.01 for (b) compared to (a), (c) and (d). Similarly,

P \ 0.01 for (d) compared to (a) and (c) which are not significantly

different from one another. (B) Bar charts showing the mean (±SEM)

sustained increases in [Ca2+]i in fura-2-loaded parotid acinar cells

following activation of CCE with normal (1.8 mM [Ca2+]o) HEPES

solution containing different [Mg2+]o (0, 1.1, 5 and 10 mM), n = 8–12

experiments taken from 5 to 10 animals. P \ 0.01 for (a) compared to

(b), (c) and (d). (Note that (a), (b), (c) and (d) represent the [Ca2+]i

signal in 0, 1.1, 5 and 10 mM [Mg2+]o, respectively at 100 s following

CCE activation)

200 Mol Cell Biochem (2008) 307:193–207

123

In this series of experiments, a number of transports

inhibitors were used to determine the mechanisms of the

ACh-evoked Mg2+ transport. Figure 10A shows samples of

original chart recordings of the effects of either 10–3 M

lidocaine, 10–3 M bumetanide and 10–4 M dinitrophenol on

[Mg2+]i in absence and presence of 10–5 M ACh. The basal

response prior to the application of each transport inhibitor

is also shown for comparison. Figure 10B shows the mean

(±SEM) increases in [Mg2+]i in magfura-2-loaded parotid

acinar cells in 1.1 mM [Mg2+]o either alone (basal) or

during perfusion with either lidocaine, amiloride, NMDG,

quinidine (all 10–3 M), 10–4 M dinitrophenol or 10–3 M

bumetanide . The results show that perfusion of magfura-

2-loaded acinar cells with either lidocaine, amiloride,

NMDG, quinidine, dinitrophenol or bumetanide resulted in

a marked and significant (P \ 0.05) elevation in [Mg2+]i

compared to the response (basal) obtained in the absence of

these transport inhibitors. The effects of bumetanide and

dinitrophenol were much more pronounced compared to

either lidocaine, amiloride or NMDG. These results suggest

that Mg2+ transport into parotid acinar cells may not be

totally dependent on extracellular Na+ since [Mg2+]i was

elevated in the presence of either NMDG or different

transport inhibitors. Figure 11A shows the effect of 10–5 M

ACh on [Mg2+]i in the presence and absence of the various

membrane transport inhibitors. The control response in the

presence of 1.1 mM [Mg2+]o is also shown for comparison.

The results show that in presence of 1.1 [Mg2+]o, ACh can

elicit a small but significant (P \ 0.05) decrease in [Mg2+]i

which was only partially blocked by either 10–3 M lido-

caine or 10–3 M amiloride, but not abolished completely. In

contrast, the ACh-evoked decrease in [Mg2+]i in the pres-

ence of either NMDG (a substitute for [Na+]o), 10–3 M

quinidine, 10–4 M dinitrophenol or 10–3 M bumetanide

was unaffected. Figure 11B shows the difference in the

ACh-evoked decrease (or changes) in [Mg2+]i in control

and in the presence of the various blockers (data taken from

Fig. 11A). The results show that the ACh-evoked decreases

in [Mg2+]i was much larger (P \ 0.01) in the presence of

bumetanide compared to the other transporter inhibitors.

Moreover, the data reveal that the decreases in [Mg2+]i

elicited by ACh were insensitive to either bumetanide,

DNP, quinidine or NMDG and only partially by either

amiloride or lidocaine.

Discussion

The interactions between cellular Mg2+ and Ca2+ have been

investigated in a number of tissues including the exocrine

Fig. 8 Basal and ACh-evoked

decreases in [Mg2+]i in

magfura-2-loaded parotid acinar

cells during perturbation of

[Mg2+]o. (A) Original chart

recordings showing the effects

of 10–5 M ACh on [Mg2+]i in

different (a) 0 mM, (b) 1.1 mM,

(c) 5 mM and (d) 10 mM)

[Mg2+]o in magfura 2-loaded

parotid acinar cells. Traces are

typical of 8–30 single cells

taken from 7 to 15 different

animals. (B) Histograms

showing the mean (±SEM)

basal [Mg2+]i in different

[Mg2+]o prior to ACh

application and the steady-state

decrease in [Mg2+]i (see

downward arrows in Fig. 8(A))

in the presence of 10–5 M Ach,

n = 8–30 experiments taken

from 7 to 15 different animals.

Note that ACh evoked

significant (*P \ 0.05)

decreases in [Mg2+]i compared

to basal for each concentration

of [Mg2+]o

Mol Cell Biochem (2008) 307:193–207 201

123

pancreas [7, 9, 10, 11], the gastric parietal cells [14] and the

heart [13]. The consensus from these previous studies is

that elevated [Mg2+]o seems to inhibit cellular Ca2+ mo-

bilisation while low [Mg2+] has the opposite effect [7, 9,

11–13] . The relationship between the Na+ and Mg2+ sig-

nalling has also been previously investigated in salivary

sublingual acini [17, 18]. Since the parotid salivary glands

are similar in structure and function to the sublingual gland

and the pancreas, and very little is known about Mg2+

homeostasis in the parotid gland, it was decided to inves-

tigate the relationship between Mg2+ and Ca2+ signalling in

this tissue. ACh was employed as the secretagogue to study

the cellular mechanism of amylase output during pertur-

bation of [Mg2+]o and [Ca2+]o. The results have

demonstrated that different concentrations of ACh can

evoke marked increases in amylase secretion from super-

fused parotid segments. A perturbation of [Mg2+]o has a

profound effect on both basal and on secretagogue-evoked

amylase secretion. Both zero and elevated [Mg2+]o can

significantly inhibit the basal as well as the secretory

effects of ACh compared to the responses obtained in

normal [Mg2+]o. The results obtained with elevated (5 and

10 mM) [Mg2+]o are in complete agreement with secreta-

gogue-evoked responses in the exocrine pancreas [7, 9, 11,

12]. In contrast to the pancreas in which zero [Mg2+]o

facilitated ACh- and CCK8-evoked amylase secretion, the

present study has revealed that zero [Mg2+]o reduced the

ACh-evoked amylase secretion from parotid segments. The

question, which now arises, is: how does a modification in

[Mg2+]o lead to an attenuation of the ACh-evoked secretory

responses in the parotid gland. Is it that [Mg2+]o is

behaving like a competitive antagonist to ACh? The

answer is obviously no, since zero [Mg2+]o produced

similar inhibition compared to elevated [Mg2+]o. The next

logical answer to the question is to understand first the

cellular mechanism of stimulus–secretion coupling process

and the physiological role of Mg2+ during cellular

regulation.

The parotid gland is innervated with autonomic nerves,

which release primarily the two main endogenous neuro-

transmitters ACh and NA upon stimulation. These in turn

activate their respective receptors (cholinergic muscarinic

for ACh and b- and a-adrenergic for NA) on parotid acinar

plasma membrane to elicit salivary protein (e.g. digestive

amylase) and fluid secretion [2, 4, 5]. ACh and a-adren-

ergic agents act via cellular Ca2+ to elicit enzyme secretion

whereas b-adrenergic activator leads to the elevation in

endogenous cyclic AMP [1, 28–30], which in turn mediates

enzyme secretion. Previously, we have shown that a

modification of [Mg2+]o had no inhibitory effect on iso-

prenaline-evoked amylase output in the parotid gland [31].

This interesting observation suggests that a perturbation of

[Mg2+]o does not seem to be associated with the stimulus–

secretion coupling pathway involving cyclic AMP. Since

ACh exerts its secretory effects via cellular Ca2+, it was

pertinent to investigate the interaction between Mg2+ and

Ca2+ during amylase output. Cytosolic Ca2+ comes mainly

from two sources, the extracellular medium and the release

from internal stores [29, 30, 32, 33–35]. If [Mg2+]o is

indeed regulating Ca2+ mobilisation then it has to control

both its influx into the cell and its release from the internal

sources. In order to test this hypothesis, [Ca2+]o was

nominally reduced during ACh-evoked amylase secretion

in different [Mg2+]o. The results presented in Fig. 2 of this

study clearly demonstrated that [Ca2+]o is required to

maintain both basal and ACh-evoked amylase secretion

and that [Mg2+]o can regulate the process probably by

controlling Ca2+ influx into the cell. Therefore, the next

logical step was to measure both basal and ACh-evoked

increase in [Ca2+]i during perturbation of [Mg2+].

The results show that basal [Ca2+]i is dependent upon

[Ca2+]o. In a nominally free [Ca2+]o and in the presence of

Fig. 9 ACh-evoked changes in [Mg2+]i in magfura-2-loaded parotid

acinar cells during perturbation of [Mg2+]o. (A) Time course changes

in [Mg2+]i during perfusion of magfura-2-loaded parotid acinar cells

with different concentrations (a) 0 mM, (b) 1.1 mM, (c) 5 mM and

(d) 10 mM [Mg2+]o and subsequently, with (e) 10–5 M ACh in

10 mM [Mg2+]o. Trace is typical of 8–10 such cells taken from 5 to 8

animals. (B) Histograms showing mean (±SEM) [Mg2+]i during

perfusion of magfura-2-loaded acinar cells with either (a) 0 mM, (b)

1.1 mM, (c) 5 mM or (d) 10 mM [Mg2+]o alone and subsequently,

with 10–5 M ACh in the presence of 10 mM [Mg2+]o, n = 8–10 such

cells taken from 5 to 8 animals. P \ 0.05 for (a) compared to (d) and

(e) compared to (d)

202 Mol Cell Biochem (2008) 307:193–207

123

1 mM EGTA, basal [Ca2+]i was significantly decreased

compared to the values obtained in the presence of 1.8 mM

[Ca2+]o. Perturbation of normal [Mg2+]o had different

effects on basal [Ca2+]i depending on the concentration of

[Mg2+]o. In zero [Mg2+]o, basal [Ca2+]i remained more or

less the same in both normal and in a nominally free

[Ca2+]o. This may be due to the release of Ca2+ from

internal stores in the absence of [Mg2+]o. Elevated [Mg2+]o

(5 and 10 mM) seem to suppress basal [Ca2+]i in both

normal and in a nominally free [Ca2+]o compared to values

obtained in 0 and 1.1 mM [Mg2+]o.

Perfusion of fura-2-loaded parotid acinar cells with a

supra-maximal dose of ACh in normal [Ca2+]o and [Mg2+]o

resulted in a large transient increase in [Ca2+]i reaching a

maximum within 10–15 s (peak phase) followed by a

decline in the Ca2+ signal reaching a steady-state plateau

above basal level after about 2–3 min (plateau phase) of

ACh application. When parotid acinar cells were perfused

with ACh in the presence of either 0, 5 or 10 mM [Mg2+]o,

there was a significant decrease in both the transient peak

and the plateau phase of the Ca2+ signal compared to the

responses obtained in 1.1 mM [Mg2+]o. The inhibitory

effect of elevated (5 and 10 mM) [Mg2+]o on the ACh-

evoked [Ca2+]i was much more pronounced compared to

the decreases obtained in zero [Mg2+]o. These results

suggest that [Mg2+]o (both zero and elevated) are regulat-

ing secretagogue-evoked cellular Ca2+ mobilisation in

parotid acinar cells.

It is now well established that the Ca2+ which is required

for the stimulus–secretion coupling process comes mainly

from two sources, one from the internal stores (e.g. endo-

plasmic reticulum) and the other from the extracellular

medium [2, 4, 5]. Since elevated [Mg2+]o can attenuate

both the transient peak and the plateau phase of the Ca2+

signals evoked by ACh, then it is tempting to suggest that

Mg2+ is exerting its inhibitory effect on Ca2+ by blocking

its influx (CCE) into the cell and its release from the

internal stores (e.g. endoplasmic reticulum).

In order to test this interesting hypothesis, parotid acinar

cells were firstly perfused with a nominally free [Ca2+]o in

presence of 1 mM EGTA and subsequently stimulated with

ACh. The cholinergic agonist evoked a small transient peak

Fig. 10 Effect of different

membrane transporters

inhibitors on [Mg2+]i in magfura

2-loaded parotid acinar cells

in the absence and presence of

10–5 M ACh. (A) Original chart

recordings showing the effect of

either (a) 10–3 M lidocaine,

(b) 10–3 M bumetanide or,

(c) 10–4 M dinitrophenol (DNP)

on [Mg2+]i in magfura 2-loaded

parotid acinar cells in the

absence and presence of 10–5 M

ACh. Traces are typical of 8–12

experiments from 5 to 6 rats.

(B) Histograms showing mean

(±SEM) basal [Mg2+]i in

1.1 mM [Mg2+]0 and the

increases in [Mg2+]i in the

presence of either 10–3 M

lidocaine (Lido), amiloride

(Amil), NMDG, quinidine

(Quin), bumetanide (Bumet) or

10–4 dinitrophenol (DNP),

n = 8–12 experiments taken

from 5 to 6 rats. P \ 0.01 for

(b), (c), (d), (e), (f) and (g)

compared to (a)

Mol Cell Biochem (2008) 307:193–207 203

123

compared to a much larger increase in normal [Ca2+]o.

Moreover, in the absence of [Ca2+]o, the plateau phase seen

normally in the presence of [Ca2+]o returned quickly to

basal level. Elevated (5 and 10 mM) [Mg2+]o significantly

attenuated the initial transient peak suggesting that it is

regulating Ca2+ release from the stores.

In contrast, elevated [Mg2+]o had no significant effect

on the plateau phase of the Ca2+ signal compared to the

response obtained in 1.1 mM [Mg2+]o in the presence

of ACh. This may be due to the fact that no Ca2+ is

entering the cell due to its absence in the extracellular

medium. Surprisingly, in zero [Mg2+]o the ACh-evoked

initial peak phase of the Ca2+ signal (see Fig. 4) was much

larger than the response (see Fig. 5) obtained in normal

(1.1 mM) [Mg2+]o. Taken together, these observations

suggest that extracellular Mg2+ is behaving like an antag-

onist for the mobilisation of cellular Ca2+. Furthermore,

when Mg2+ is absent from the extracellular medium, more

Ca2+ seems to be released from internal stores. Previous

studies have suggested that Mg2+ is a natural antagonist for

Ca2+ [7, 11].

The results so far have clearly indicated that Mg2+ can

regulate Ca2+ release from the internal stores and more-

over, the divalent cation may also control Ca2+ influx into

the cell during the CCE process. The experiments pre-

sented in Figs. 6 and 7 were done to test this interesting

hypothesis [29, 30]. The results have demonstrated that

normal (1.1 mM) [Mg2+]o can rapidly activate the CCE

compared to the delayed activation times in 0, 5 and

10 mM [Mg2+]o. In addition, the data have also shown that

maximal elevation in [Ca2+]i in parotid acinar cells occurs

in zero [Mg2+]o compared to significantly less Ca2+ entry

in the presence of elevated [Mg2+]o, especially in the

presence of 5 and 10 mM [Mg2+]o. The results of this

study are in total agreement with the data obtained in both

the pancreas and parotid gland in which elevated [Mg2+]o

inhibited both Ca2+ release from intracellular stores and

Ca2+ influx from extracellular medium [7, 9, 12, 25, 36].

Mg2+ can exert its effect on Ca2+ transport either directly

or via its effects on Mg2+-dependent enzymes in the cell

[6, 7, 16]. Therefore, the most logical approach was to

measure [Mg2+]i in parotid acinar cells and characterise its

transport mechanism.

The results show that perfusion of single magfura-2-

loaded parotid acinar cells with HEPES solution con-

taining different concentrations of [Mg2+]o resulted in a

gradual increase in [Mg2+]i. Stimulation of acinar cell

with ACh in different [Mg2+]o resulted in a gradual

decrease in [Mg2+]i reaching a plateau phase within

5–8 min. These findings suggest that ACh can mobilise

cellular Mg2+. The decrease in the Mg2+ signal may be

due to the fact that Mg2+ is either leaving the cell or it is

entering intracellular stores. If it is exiting the cell, then it

has to move against a gradient since the secretagogue-

evoked decrease was obtained in elevated (5 and 10 mM)

[Mg2+]o suggesting that this movement is dependent upon

energy. These results of ACh are in agreement with the

data obtained in previous studies employing pancreatic

acinar cells [7, 10, 11, 24] as well as salivary sublingual

acini [17, 18].

It has been previously demonstrated, at least in both

sublingual and pancreatic acinar cells, that the secreta-

gogue-evoked decreases in [Mg2+]i were due to Mg2+

leaving the cell and that the efflux of Mg2+ was dependent

upon extracellular Na+ [17, 18, 36]. Some of these studies

were also done employing the techniques of atomic

absorbance spectroscopy and magfura-2 tetrapotassium salt

Fig. 11 Changes in [Mg2+]i in magfura-2-loaded parotid acinar cells

in the absence and presence of different membrane transporter

inhibitors. (A) Histograms showing mean (±SEM) changes in [Mg2+]i

following perfusion of magfura-2 acinar cells with either 1.1 mM

[Mg2+]o HBSS solution alone or with HBSS solution containing either

10–3 M lidocaine (Lido), 10–3 M amiloride (Amil), NMDG, 10–3 M

quinidine (Quin), 10–3 M bumetanide (Bumet) or 10–4 M dinitrophe-

nol (DNP) in the absence and presence of ACh 10–5 M, n = 8–12

experiments taken from 5 to 6 rats. (B) Histograms showing mean

(±SEM) of the ACh-evoked decreases in [Mg2+]i, either alone (a) or

in the presence of each inhibitor (b–g), n = 8–12 experiments.

P \ 0.05 for (a) compared to (d), (e), (f) and (g). Note that the data

are obtained from A[J1]

204 Mol Cell Biochem (2008) 307:193–207

123

to measure Mg2+ efflux [7, 11, 22]. However, in the parotid

acinar cells no previous study has attempted to characterise

either basal or secretagogue-evoked Mg2+ transport. The

results of this study have also shown that either lidocaine

(a Na+ channel blocker), amiloride (an inhibitor of Na+/H+

exchanger), NMDG (a substitute for [Na+]o), quinidine (an

inhibitor of the Na+/Mg2+ antiport), dinitrophenol (an

inhibitor of ATP) and bumetanide (an inhibitor of the

Na+:K+:Cl- co-transporter) can all increase [Mg2+]i with the

same order of potency, with lidocaine eliciting the least

increase and bumetanide the maximal increase. In the

presence of any of these inhibitors, ACh evoked a decrease

in [Mg2+]i. The magnitude of the ACh-evoked decrease in

[Mg2+]i was much bigger in the presence of bumetanide

and much more smaller in the presence of either lidocaine

or amiloride. The decrease in [Mg2+]i in response to ACh

may be due to the fact that Mg2+ is leaving the cell (efflux)

similarly to the results obtained in pancreatic acinar cells

during ACh or CCK-8 stimulation [10, 11, 19, 24] or it is

sequestered into intracellular stores as in the sublingual

gland [17, 18]. Since parotid acinar cells are similar to

sublingual and pancreatic acinar cells, it is tempting to

suggest that ACh is indeed stimulating Mg2+ efflux and

probably also its influx into organelles within the cell. If

this is the case, then this ‘‘ACh-induced Mg2+ efflux’’ is

insensitive to either Na+ removal (substituting it for

NMDG) or to either quinidine, dinitrophenol or bumeta-

nide but only partial sensitive to lidocaine and amiloride.

Taken together, the results suggest that the ‘‘ACh-evoked

Mg2+ efflux’’ may partially be associated with the sodium

channel activity since lidocaine, the local anaesthetic

which is known to inhibit the Na+ channel activity, can

decrease the response to ACh. In addition, the ‘‘ACh-

evoked Mg2+ efflux’’ may also be associated with the Na+/

H+ antiport. Previous studies employing salivary sublingual

gland have demonstrated that secretagogue-evoked Mg2+

transport was mediated by a Na+ dependent-pathway

[17, 18].

On the other hand, the transport inhibitors themselves

and sodium substitution (with NMDG) can result in sig-

nificant elevations in [Mg2+]i compared to the basal value

similar to that obtained in the sublingual acini [17, 18].

These transport inhibitors may exert their effects in

increasing cellular Mg2+ via a different number of mech-

anisms. They may act by either facilitating Mg2+ release

from intracellular stores [17, 18], preventing its efflux from

the cytosol or enhancing its influx from the extracellular

medium. Since [Mg2+]i rises in the presence of NMDG, it

is tempting to suggest that cytosolic Mg2+ elevation is not

sensitive to extracellular Na+ at least in parotid acinar cells.

It is particularly noteworthy that in sublingual acinar cells

that Mg2+ is released from an intracellular pool and this is

mediated by a sodium-dependent magnesium transport

mechanism [17, 18]. Like NMDG, the same is true for

bumetanide. In the presence of the loop diuretic, there was

a marked elevation in [Mg2+]i. Bumetanide is known to

inhibit the Na+:K+:Cl- co-transport. The results with DNP

indicate that the rise in [Mg2+]i may be dependent on an

energy process since DNP is known to inhibit ATP pro-

duction. Another possible explanation for the effect of

DNP is that basal Mg2+ efflux is dependent upon ATP and

Fig. 12 Schematic model illustrating the interaction between Mg2+

and Ca2+-signalling during amylase secretion in parotid acinar cells in

response to ACh. Following stimulation, the secretagogue evoked an

increase in [Ca2+]i from the endoplasmic reticulum (ER) which in turn

activates CCE leading to the stimulation of calmodulin (CD), Ca2+-

CD activates the phosphorylation of regulatory proteins on the

salivary protein granules resulting in the influx of ions and water and

subsequent swelling of the granules. The granules then migrate

towards the luminal pole where they dock and fuse with the luminal

membrane to bring about exocytosis and secretion. It is proposed that

Mg2+ can regulate the metabolism of IP3, Ca2+-ATPase pumps

(SERCA and PMCA) in the endoplasmic reticulum (SERCA) and

plasmic membrane (PCMA), respectively, leading to Ca2+ release

from the ER, and Ca2+ influx from the extracellular medium. High

[Mg2+]o, and subsequently high [Mg2+]i, seems to attenuate Ca2+

release from the ER and its entry into the cell, whereas low [Mg2+]o

and subsequently, [Mg2+]i; has effects on enzymes which regulate

cellular Ca2+ homeostasis. PLC = phospholipase C; IP3 = inositol

trisphosphate; PIP2 = phosphatidyl inositol biphosphate; DG = diac-

ylglycerol; CCE = capacitative calcium entry; IP3R = IP3 receptor;

RyR = ryanodine receptor

Mol Cell Biochem (2008) 307:193–207 205

123

once inhibited with DNP, this resulted in an elevation of

[Mg2+]i. Similarly, it can be argued that Mg2+ uptake into

internal stores is ATP-dependent and this process could

then be inhibited by DNP, resulting in a rise in [Mg2+]i.

Furthermore, quinidine can also elevate [Mg2+]i and this

substance is known to inhibit the Na+/Mg2+ antiport. Like

NMDG, quinidine would inhibit Na+ influx into the cell

hereby facilitating cellular Mg2+ elevation. Like the sub-

lingual gland, further experiments are required to

characterise the release of Mg2+ from internal stores and its

transport from the cytoplasm into the intracellular organ-

elles in the parotid acinar cells.

The schematic model in Fig. 12 summarises the inter-

action between Ca2+ and Mg2+ in parotid acinar cell during

stimulation with ACh. It is proposed that a perturbation of

[Mg2+]o can attenuate ACh-evoked Ca2+ mobilisation, both

its influx or efflux from extracellular medium and its

release or uptake from the endoplasmic reticulum via its

effect on enzymes which regulate Ca2+ homeostasis. The

Ca2+ in turn interacts with calmodulin to phosphorylate

parotid cell zymogen granules to mediate amylase secre-

tion. The results also show ACh can elicit a decrease in

[Mg2+]i which is insensitive to a number of membrane

transport inhibitors and strongly associated with the stim-

ulus–secretion coupling events. In agreement with previous

findings, it seems that the ACh-evoked Mg2+ mobilisation

occurs in a way that is completely anti-parallel to the rapid

Ca2+ signalling in acinar cells [11]. Finally, the regulation

of basal levels of [Mg2+]i is tightly controlled and it may be

dependent upon the presence of extracellular Na+. These

observations indicate a regulatory role of Mg2+ in the

physiology as well as the pathophysiology of the salivary

parotid gland.

References

1. Baum BJ (1987) Neurotransmitter control of secretion. J Dent

Res 66:628–632

2. Baum BJ (1993) Principles of salivary secretion. In: Saliva as a

diagnostic fluid. Eds. Ann NY Acad. Sci; 694. New York.

pp 17–23

3. Baum BJ, Ambudkar IS (1988) Regulation of calcium handling

by rat parotid acinar cells. Mol Cell Biochem 82:67–73

4. Petersen OH, Gallacher DV (1988) Electrophysiology of pan-

creatic and salivary acinar cells. Ann Rev Physiol 50:65–80

5. Ambudkar IS (2000) Regulation of calcium in salivary gland

secretion. Crit Rev Oral Biol Med 11:4–25

6. Birch NJ (1993) Magnesium and Cell. Academic Press, London

7. Yago MD, Manas M, Singh J (2000) Intracellular free magne-

sium regulation and transport in epithelial secretory cells. Front

Biosci 5:602–619

8. Reinhart RA (1988) Magnesium metabolism. A review with

special reference to the relationship between intracellular content

and serum levels. Arch Intern Med 148:2415–2420

9. Francis LJ, Lennard R, Singh J (1990) Mechanism of action of

magnesium on acetylcholine evoked secretory responses in the

isolated rat pancreas. Exp Physiol 75:669–680

10. Lennard R, Singh J (1992) Effects of secretagogues on intracel-

lular free calcium and magnesium concentrations in rat pancreatic

acinar cells. Gen Pharmacol 23:903–908

11. Mooren FC, Turi S, Gunzel D et al (2001) Calcium–magnesium

interactions in pancreatic acinar cells. FASEB J 15:659–672

12. Wisdom DM, Salido GM, Baldwin LM, Singh J (1996) The role

of magnesium in regulating CCK-8 evoked secretory responses in

the exocrine rat pancreas. Moll Cell Biochem 154:123–132

13. Chakraborti S, Chakraborti T, Mandal M et al (2002) Protective

role of magnesium in cardiovascular diseases: a review. Mol Cell

Biochem 238:163–179

14. Mooren FC, Geada MR, Singh J et al (1997) Effects of extra-

cellular Mg2+ concentration on intracellular signalling and acid

secretion in rat gastric parietal cells. Biochem Biophys Acta

1538:279–288

15. Altura BM, Altura BT (1985) New perspectives on the roles of

magnesium in the pathophysiology of the cardiovascular system.

I Clinical aspects. Magnesium 4:226–244

16. Flatman P W (1984) Magnesium transport across cell mem-

branes. J Membr Biol 80:1–14

17. Zhang GH, Melvin JE (2005) Regulation of extracellular Na+ of

cytosolic Mg2+ concentration in Mg2+-loaded rat sublingual acini.

FEBS Lett 371:52–56

18. Zhang GH, Melvin JE (2006) Na+-dependent release of Mg2+

from an internal pool in rat sublingual mucous acini. J Biol Chem

271:29067–29072

19. Gonzalez A, Pariente JA, Salido GM et al (1997) Intracellular pH

and calcium signalling in rat pancreatic acinar cells. Pflugers

Arch—Eur J Physiol 434:609–614

20. Rinderknecht J, Marbach EP (1970) A new automated method for

the determination of serum a-amylase. Clin Chem Acta 29:107–

110

21. Streb H, Irvine RF, Berridge MJ et al (1983) Release of calcium

from a non-mitochondrial intracellular store in pancreatic acinar

cells by inositol 1,4,5 trisphosphate. Nature 306:67–68

22. Herzog V, Sies H, Miller F (1976) Exocytosis in secretory cells of

rat lachrymal gland. J Cell Biol 70:692–706

23. Gonzalez A, Camello PJ, Pariente JA et al (1997) Free cytosolic

calcium levels modify intracellular pH in rat acini. Biochem

Biophys Res Commun 230:652–656

24. Lennard R, Singh J (1991) Secretagogue-evoked changes in

intracellular free magnesium concentrations in rat pancreatic

acinar cells. J Physiol 435:483–492

25. Pariente JA, Camello C, Camello PJ et al (2000) Release of

calcium from mitochondrial and non-mitochondrial intracellular

stores in mouse pancreatic acinar cells by hydrogen peroxide.

J Membr Biol 179:27–35

26. Pariente JA, Redondo PC, Granados MP et al (2005) Calcium

signalling in non-excitable cells. Eur Citz Qual Life 1:29–43

27. Putney Jr JW (1988) The role of the phosphoinositide metabo-

lism in signal transduction in secretory cells. J Exp Biol

139:135–158

28. Putney Jr JW, Bird GS (1993) The inositol-phosphate calcium

signalling system in non excitable cells. Endocrine Rev 14:610–

631

29. Putney Jr JW (1986) Identification of cellular activation mecha-

nisms associated with salivary secretion. Ann Rev Physiol

30:209–245

30. Putney Jr JW (1986) A model for receptor-regulated Ca2+ entry.

Cell Calcium 7:1–12

31. Mata AD (2003) Influence of magnesium on salivary gland

secretion: physiological and pathophysiological studies. PhD

206 Mol Cell Biochem (2008) 307:193–207

123

Thesis. University of Central Lancashire, Preston, United

Kingdom

32. Berridge M, Lipp P, Bootman M (1999) Calcium signalling.

Bioassays 21:R157–R159

33. Berridge MJ (1997) Elementary and global aspects of calcium

signalling. J Exp Biol 200:315–319

34. Carafoli E (1987) Intracellular calcium homeostasis. Annu Rev

Biochem 56:395–433

35. Petersen OH (1992) Stimulus–secretion coupling: cytoplasmic

Ca2+ signals and control of ion channels in exocrine acinar cells.

J Physiol 448:1–51

36. Yago MD, Mata AD, Manas M et al (2002) Effect of extracellular

magnesium on nerve mediated and acetylcholine-evoked secre-

tory responses in the isolated rat parotid gland. Exp Physiol

87:321–326

Mol Cell Biochem (2008) 307:193–207 207

123